Depressive disorders are among the most prevalent and poorly managed problems in American primary care practice. As an initial response to this situation, the Agency for Health Care Policy and Research commissioned an expert panel to create Clinical Practice Guidelines for the detection and management of major depression in primary care patients. To understand their potential impact, the field needs to determine how a concerted effort to disseminate the Guidelines will affect the process, outcomes and costs of routine clinical care. To address this need, we propose a project with three specific aims. First, we will test how an intervention targeting both primary care physicians and depressed patients affects the probability that patients with major depression will be detected, and the probability that depressed patients receive a recommendation for and complete an acute phase course of Guideline-concordant care. Second using an intent to treat analysis, we will measure the effect of the intervention on patient outcomes and overall health care expenditures at six month follow-up. Third, because features of the intervention besides Guideline treatment itself may improve outcomes, we propose to conduct a quasi-experimental replication of the study's main findings to compare outcomes (and expenditures) in all subjects who do and do not complete acute-stage Guideline-concordant care using selection models to control for unobserved differences between the two groups. We propose to use an experimental design in which 24 family physicians are matched on their reported depression practice patterns before one member of each pair is randomized to the experimental and the other to the comparison group. Each practice will recruit a cohort of 20 patients meeting criteria for current major depression (total n = 480) using a protocol we have successfully employed in previous research. The 12 physicians randomized to the intervention condition will participate in a series of educational conference calls, receive reminders about Guideline recommendations for the patient's treatment of choice, and consult with a psychiatrist on treatment-resistant patients. Depressed patients identified through a two- stage screening process, will receive a complementary patient education intervention. Patient outcomes will be measured using a self-administered instrument we have validated to measure the outcomes of care for major depression in primary care settings 6 months following the index visit. Expenditures for care will be measured using a methodology we pioneered to estimate overall health care expenditures. We have also designed the study to be able to conduct exploratory research on how the intervention affects physician knowledge and attitudes about the clinical utility of the Guidelines. The project will be conducted in the Ambulatory Practice Sentinel Network (ASPN), whose patients and presenting problems are highly comparable to those of the general U.S. population seeking care from family physicians. Thus, this application will provide generalizable evidence about how a potentially disseminable intervention to promote Guideline-concordant treatment for major depression in primary care patients affects the process, outcomes and costs of care in sociodemographically and organizationally diverse community-based practices across the country.